Martin Gauch
@gauchm.bsky.social
Deep learning & earth science @ Google Research
Back by popular demand: At #EGU26 we'll organize another BUGS session: Blunders, Unexpected Glitches, and Surprises!
Submit abstracts on ideas that seemed great but didn't work, errors and bugs that led to new insights (or funny stories), or any other unexpected results.
www.egu26.eu/session/56997
Submit abstracts on ideas that seemed great but didn't work, errors and bugs that led to new insights (or funny stories), or any other unexpected results.
www.egu26.eu/session/56997
December 3, 2025 at 1:58 PM
Back by popular demand: At #EGU26 we'll organize another BUGS session: Blunders, Unexpected Glitches, and Surprises!
Submit abstracts on ideas that seemed great but didn't work, errors and bugs that led to new insights (or funny stories), or any other unexpected results.
www.egu26.eu/session/56997
Submit abstracts on ideas that seemed great but didn't work, errors and bugs that led to new insights (or funny stories), or any other unexpected results.
www.egu26.eu/session/56997
It's (finally) published: hess.copernicus.org/articles/29/...
Looking forward to all the different ways the title will be messed up by indexing tools!
Looking forward to all the different ways the title will be messed up by indexing tools!
November 13, 2025 at 9:30 AM
It's (finally) published: hess.copernicus.org/articles/29/...
Looking forward to all the different ways the title will be messed up by indexing tools!
Looking forward to all the different ways the title will be messed up by indexing tools!
Reposted by Martin Gauch
Congratulations to Frederik Kratzert on winning this year's Arne Richter Award for outstanding research by an early career scientist. Fantastic presentation at #EGU25 this afternoon!
May 1, 2025 at 12:18 PM
Congratulations to Frederik Kratzert on winning this year's Arne Richter Award for outstanding research by an early career scientist. Fantastic presentation at #EGU25 this afternoon!
Reposted by Martin Gauch
Now on HESSD for open discussion: egusphere.copernicus.org/preprints/20...
They even let us keep the paper title (for now?!) 🙄
They even let us keep the paper title (for now?!) 🙄
April 7, 2025 at 11:48 AM
Now on HESSD for open discussion: egusphere.copernicus.org/preprints/20...
They even let us keep the paper title (for now?!) 🙄
They even let us keep the paper title (for now?!) 🙄
NeuralHydrology just got a little better, especially if you're building custom models :)
After a long time, we published a new version of NeuralHydrology, see github.com/neuralhydrol...
Most relevant:
- Added all code related to our recent preprint eartharxiv.org/repository/v...
- Finally make the Dataset return each feature separately, rather than stacked into one tensor.
Most relevant:
- Added all code related to our recent preprint eartharxiv.org/repository/v...
- Finally make the Dataset return each feature separately, rather than stacked into one tensor.
Release v1.12.0 · neuralhydrology/neuralhydrology
Added
Added input replacing, masked mean, and attention support to InputLayer. This allows the models to handle NaN inputs. More details in the preprint How to deal w___ missing input data. Use na...
github.com
March 18, 2025 at 9:00 AM
NeuralHydrology just got a little better, especially if you're building custom models :)
Starting on bsky with a new preprint: "How to deal w___ missing input data"
doi.org/10.31223/X50...
Missing input data is a very common challenge in deep learning for hydrology: weather providers have outages, some data products start later than others, some only exist for certain regions, etc.
doi.org/10.31223/X50...
Missing input data is a very common challenge in deep learning for hydrology: weather providers have outages, some data products start later than others, some only exist for certain regions, etc.
March 15, 2025 at 12:50 PM
Starting on bsky with a new preprint: "How to deal w___ missing input data"
doi.org/10.31223/X50...
Missing input data is a very common challenge in deep learning for hydrology: weather providers have outages, some data products start later than others, some only exist for certain regions, etc.
doi.org/10.31223/X50...
Missing input data is a very common challenge in deep learning for hydrology: weather providers have outages, some data products start later than others, some only exist for certain regions, etc.